Abstract
This article discusses the generation and interpretation of spatial references in spoken language interfaces. A dialogue-based user guide for a complex machine is presented, utilising a 3D model of the machine and spatial-based reasoning. Two issues are discussed. The first issue is the architecture of the user guide supporting both 3D graphical presentations and graphics-based reasoning within the dialogue process. The second issue concerns algorithms for spatial reasoning used in the system. The following techniques are discussed: the use of functional hierarchy, reduction of dimension, and the use of zones of applicability for spatial references. It is shown that knowledge of the functionality and hierarchical relations between the machine components can be used to enhance reasoning.
Keywords
Feature modeling, collaborative modeling, web-based modeling, graphical interaction